2016 International Conference on Big Data and Smart Computing (BigComp) (2016)
Hong Kong, China
Jan. 18, 2016 to Jan. 20, 2016
Zae Myung Kim , School of Computing, KAIST, Korea
Young-Seob Jeong , School of Computing, KAIST, Korea
Ho-Jin Choi , School of Computing, KAIST, Korea
With the exponential growth in data, we often find ourselves struggling to deal with information overload. Techniques such as timeline summarization tackle this problem by generating short summaries for each time stamp on a timeline. However, we argue that rather than reading a set of blocks of texts, it is easier and quicker for a reader to observe dynamically changing relations between important entities that are illustrated through graphs. It enables her/him to grasp the major events of the whole story quickly. We develop an experimental system to evaluate on the 2014 FIFA World Cup Brazil news articles, and show that the result covers major events of the World Cup and can be understood in a short span of time.
Visualization, Feature extraction, Sparse matrices, Data mining, Information retrieval, Computational linguistics, Optimization
Zae Myung Kim, Y. Jeong and Ho-Jin Choi, "Understanding news stories through SVO triplets," 2016 International Conference on Big Data and Smart Computing (BigComp)(BIGCOMP), Hong Kong, China, 2016, pp. 498-501.